EEVblog #828 - Siglent SPD3303X precision Lab PSU Teardown

[EEVblog]

Inside the Siglent SPD3303X Precision Programmable DC Lab Power Supply.
http://amzn.to/1m9Knp2
UPDATE: There is an even cheaper version, the SPD3303X-E which has 10mV/10mA resolution.
http://amzn.to/1jNsQRL

[brutester]

At around 13:20 I would say that those diodes are for system supply for CPU, DAC, etc, not 3rd output

[EEVblog]

At around 13:20 I would say that those diodes are for system supply for CPU, DAC, etc, not 3rd output

Nope, keep watching, the CPU etc has it's own bridge and big filter caps on the front panel board. The bridge gets damn hot too.

[Opticalworm]

Good news Dave, linux is probably not an option for this supply. The STM32F207VE is a cortext M3 with around 512KB of memory.

My money is on an RTOS. 

-Ron

[rs20]

26:30 -- If you have the output physically shorted, the voltage you choose is irrelevant. It's in constant current mode. So there's no point in settings it to 1V or whatever to make it maximally stressful, you've already achieved that with the shorted outputs.
29:30 -- Is this oscillation caused by the dummy load, or by the PSU (or by the combined feedback loop)? Put another way, would we see this same oscillation with a real, physical 50W 1 ohm resistor? I'm happy to buy you the resistor if you're willing to test it 

[EEVblog]

Good news Dave, linux is probably not an option for this supply. The STM32F207VE is a cortext M3 with around 512KB of memory.

Ah, ok, I didn't check the processor specs.

[dr.diesel]

Mmmm,

 - Deceptive Corporate practices
 - Non standard width banana plugs
 - Crap quality banana plugs
 - Shit caps

Seriously, those banana plugs?  Entire instrument is a complete fail.

 

[FrankBuss]

I didn't notice the banana plugs problem, looks my Siglent SPD3303D has the same problem. With the cables I'm usually using it is no problem, but they are loose, too, with some old connectors with the slit instead of the bulges.

I've done a short review of the SPD3303D, see here. There was a major problem with the fixed voltage output, but I could patch it with the help of the Siglent support. Would be interesting if the SPD3303X has already fixed this problem. And I like the new voltage adjustment concept. Too bad that they probably won't change it for the SPD3303D anymore. Still a very nice power supply unit for the price.

[madires]

- Deceptive Corporate practices
 - Non standard width banana plugs
 - Crap quality banana plugs
 - Shit caps

And the shorted turn of the toroidal transformer. The mounting screw doesn't seem to be isolated. Rust, again. The bodged ethernet. Build quality is the same like my Mastech clone. I wouldn't pay more than €300 for that Lab PSU.

[thomastheo]

If they managed to overlook the fact that their binding posts don't work it really makes me wonder what else escaped their attention. That is an immediate deal breaker, i can't think of anything more basic in terms of requirements than adequate binding posts.

[WN1X]

Those binding posts are absolute crap. How difficult is it to get the spacing right?!?

Sorry, but my next lab supply will most likely be the Rigol DP832.

[madires]

If they managed to overlook the fact that their binding posts don't work it really makes me wonder what else escaped their attention. That is an immediate deal breaker, i can't think of anything more basic in terms of requirements than adequate binding posts.

Absolutely! Siglent should rework all those details already mentioned. After seeing the teardown I doubt anyone would buy that PSU. BTW, there's a little brother "-E" with a 10mV/mA resolution for €370 (SPD3303X is €509). Same issues, I'd guess.

[c4757p]

The mains wire rubbing against the sharp chassis edge is 100% inexcusable. It already has scratch marks! That is not something you can get away with - over time that will rub through. I'd have stopped the teardown right there and sent the thing back immediately as a fire/shock hazard.

What the hell are Siglent playing at? That's not just cheap and crappy, it's downright dangerous. The engineer who did that should be fired.

[Artlav]

Tangential question - what in the world would you need a power supply with such a precision for?

[Deathwish]

Whats with the Siglent / Rigol comparisons all the time, it turns everything into a two horse race.

The fact this one sucks worse than a two dollar hooker says it all. But at least compare it to some other brand than Rigol.

[Robaroni]

Thanks Dave, nice tear down as usual.

OK, the output jacks, sure they should have given us better quality - easy fix. Spacing?  Not a game changer, I use supplies everyday, rarely if ever need the standard spacing but I guess they could fix it in an update but the cost would probably be high as the front panel would have to go as would the some of the internals but, again, I can live with it.

Small details like the unprotected leads, easy fix. The caps? Well the only way to know if the quality is there is to push them on a test jig. I wouldn't trash them before that was done, it's not fair to Siglent, after all we don't know what their testing of components was.

The screen, OK easy software fixes, not a big deal.

I like the size over the DP832 series, the less room these things take on  bench the better. The Rigol is a big supply about 43cm (17") in depth and I want a triple supply to limit space!

My gripe is the 5 volt supply without current control. Now that can burn up some costly chips and it's something I use on a daily basis. Think about it, you have a breadboard full of components and you accidentally short outputs on a Maxim chip that goes for 10 bucks. Ouch!! I read spec sheets so when I first turn on a prototype I have current limit right above specs. And what happens if you short the 5v supply? That's one you should have tried Dave. Does it blow the soldered fuse inside the case? Now that concerns me.

Size, the 1ma resolution, those are good things I like, it just needs some polish, clean up the inside of the case problems, support the TO220's and do something about the 5v current problem and I'm interested.

Rob

[FrankBuss]

And what happens if you short the 5v supply? That's one you should have tried Dave. Does it blow the soldered fuse inside the case? Now that concerns me.

I guess it uses the same circuit as the SPD3303D which I've tested and of course it doesn't blow the fuse if you short the 5V supply. Probably they have even fixed the nasty turn-on spike problem on this output.

[mikron]

There is a screw missing ...
Maybe Dave removed the PCB to check clearances?

[free_electron]

Non standard pitch on the banana jacks ( for the bnc adapter ) -> immediate disqualification in my book. no matter how good the thing is otherwise.

Sense wires ?

[lukier]

Sense wires ?

If it claims to be a precision power supply I think it should have sense inputs.

On the other hand there are plenty of much more expensive PSUs in the similar category (universal multi output bench type) that don't have this feature. Think HP E3631A and Rigol's clone - DP1308A. I think some other models from HP E3xxx series had them though.

[Robaroni]

And what happens if you short the 5v supply? That's one you should have tried Dave. Does it blow the soldered fuse inside the case? Now that concerns me.

I guess it uses the same circuit as the SPD3303D which I've tested and of course it doesn't blow the fuse if you short the 5V supply. Probably they have even fixed the nasty turn-on spike problem on this output.

Thanks Frank,
What's the current capacity of that supply? Like I said, I don't like a supply in 2015 that doesn't have a current regulated output. Unfamiliar caps I can live with if the quality is there and I think Dave should be less harsh on his condemnations.

Basically it's a dual supply, not a triple, the DP832 is a real triple whatever other warts it might have.
Rob

[Rasz]

 

is there really ~1000microF output capacitance?

Cant wait for excuses, 'we let flesh college graduate design this supply and source bottom of the barrel components'?

[Monkeh]

That thing is a mess.

And for a brand new product, it's interesting how the chassis is nearly identical to one I've had for years. .. and the heatsinks, and actually, the whole power PCB..

Just a quick, cheap clone with a UI and some awful binding posts. Poor show, Siglent. Very poor.

[nrxnrx]

Does the display really turn off after some "inactivity"? Even if it allows you to turn that off, they made it the default?

I recently bought some cables (4mm banana) listed as "unbranded" but very cheap. I can't use them, but I guess they'd fit this 539USD power supply.

--
Mihai

[Kevin.D]

The banana plugs are poor mistake, always need good quality O/P sockets considering how often they are used.
Any high Voltage Mains/primary wiring and low Voltage Secondary wiring are really best kept completely seperate, with the high Voltage mains wiring having also an outer sheeth protection.
Any serious bench supply needs remote sensing, I use feature often.
Otherwise internal/overall construction looks fairly well done and tidy. Needs a proper review doing with load stepping.

is there really ~1000microF output capacitance?

Cant wait for excuses, 'we let flesh college graduate design this supply and source bottom of the barrel components'?

No they are 470Uf, see 17:33 in the vid, Which is still on the biggish side . Rigol dp832 is 1000uF I believe.

Note at ~16:40 in the video the row of large brown ac caps (is that a 400v rating ?). These are probably the Common Mode output filter caps , these are important ( unless using an transformer with very low inter-winding capacitance) to keep common mode noise to minimum. Something that rigol dp832 completely forgot about apparently. 


Regards